2014 IEEE International Conference on Big Data (Big Data) 2014
DOI: 10.1109/bigdata.2014.7004382
|View full text |Cite
|
Sign up to set email alerts
|

Privacy-aware filter-based feature selection

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

0
5
0

Year Published

2015
2015
2023
2023

Publication Types

Select...
4
2
2

Relationship

0
8

Authors

Journals

citations
Cited by 10 publications
(5 citation statements)
references
References 14 publications
0
5
0
Order By: Relevance
“…For correlated or otherwise related variables (often referred to as proxies or quasi-identifiers), much of the literature assumes a priori knowledge of the set of quasi-identifiers [105], or seeks to discover them on a case-by-case basis (e.g. [206,139]), and moves towards a notion of privacy preserving data mining as introduced by [8]. [123] also discuss the notion of proxy groups, a set of "similar" instances of the data that could correspond to a protected group (e.g.…”
Section: Sensitive and Protected Variables And (Un)privileged Groupsmentioning
confidence: 99%
“…For correlated or otherwise related variables (often referred to as proxies or quasi-identifiers), much of the literature assumes a priori knowledge of the set of quasi-identifiers [105], or seeks to discover them on a case-by-case basis (e.g. [206,139]), and moves towards a notion of privacy preserving data mining as introduced by [8]. [123] also discuss the notion of proxy groups, a set of "similar" instances of the data that could correspond to a protected group (e.g.…”
Section: Sensitive and Protected Variables And (Un)privileged Groupsmentioning
confidence: 99%
“…Jafer et al [20] have proposed a privacy-aware filter-based feature selection that probes the inter-correlation of features to remove quasi-identifier (QI) features. In their paper, the authors introduce a system which contains two separate blocks: one for evaluating features, and the other one for controlling the privacy aspects of feature selection.…”
Section: Related Workmentioning
confidence: 99%
“…By referring to the controlling values of the Correlation Block, features are selected or discarded. We have adopted Figure 4 from [20] to illustrate the proposed system. After evaluating features using e, features are sorted in descending order list.…”
Section: Related Workmentioning
confidence: 99%
“…While there is research on identifying the most informative non-sensitive attributes for a confidential attribute [21], such research aims at producing sanitized public data sets. Thus, in such a case, one authority globally decides which attributes are sensitive and which ones are not, and records of all individuals are treated the same.…”
Section: Personalized Protective Analyticsmentioning
confidence: 99%